converter

Abstract

To provide a more excellent decoupling technology capable of suppressing interference between two ports, in a TAB converter.SOLUTION: Provided is a TAB converter that comprises three ports each having a full-bridge circuit. A control circuit 90 executes a step of calculating a control value by using an approximate current waveform of a port current to a target current. The approximate current waveform of the target current is derived from an approximate voltage waveform of a port voltage outputted from each full bridge circuit. The approximate voltage waveform is a continuous function obtained by superposing a fundamental wave and a high harmonic wave. Thereby, interference between two output ports can be more suppressed compared with a case where an approximate model approximated only by the conventional fundamental wave is used.SELECTED DRAWING: Figure 2

Description

[Technical Field]

[0001] This invention relates to a converter having three ports. [Background technology]

[0002] Conventionally, dual active bridge (DAB) converters, which have full bridge circuits on both the input and output sides, are known. Furthermore, triple active bridge (TAB) converters, which add one more port to the DAB converter, are being researched, and by applying a TAB converter, it becomes possible to transmit power bidirectionally between the three ports.

[0003] A TAB converter is a converter in which three active bridge circuits are connected via a three-winding transformer and inductors, and power is transmitted by utilizing the phase shift of each bridge circuit. Because a TAB converter has the characteristic that the sum of the input and output power of each port is zero, the input and output current of the remaining port is determined by controlling the current of two ports. On the other hand, there is a problem that interference occurs between the two ports being controlled. A conventional method for suppressing interference between output ports is described, for example, in Non-Patent Document 1. [Prior art documents] [Non-patent literature]

[0004] [Non-Patent Document 1] Takanobu Ohno and Shinichi Hoshi, "Experimental Study of Inter-Port Interference in a Triple Active Bridge Converter," Transactions of the Institute of Electrical Engineers of Japan, Vol. 139, No. 7, pp. 631-636, 2019. [Overview of the project] [Problems that the invention aims to solve]

[0005] The interference suppression method described in Non-Patent Document 1 approximates the voltage of a full-bridge circuit driven by a square wave with the fundamental wave, thereby mathematically modeling the current at each port of the TAB converter as a continuous function. Decoupling is then performed by inserting a compensator that cancels out the interference between ports in that model. While this method can achieve a certain degree of decoupling, depending on the application of the output power, a method for achieving better decoupling is required.

[0006] Here, the "fundamental wave" is the lowest frequency sine wave obtained when a waveform with certain frequency components is expanded into a Fourier series. Furthermore, the sine wave with a frequency n times that of the fundamental wave, obtained when a waveform with certain frequency components is expanded into a Fourier series, is called the "nth harmonic."

[0007] Therefore, the present invention aims to provide a superior technology for decoupling in a TAB converter that suppresses interference between two ports. [Means for solving the problem]

[0008] The first invention of the present application is a triple active bridge converter having three ports, comprising: a transformer having a first winding, a second winding, a third winding, and a core that magnetically couples the first, second, and third windings to each other; a first port connected to the first winding, a second port connected to the second winding, a third port connected to the third winding, and a control circuit for switching control of the switching elements of the first, second, and third ports, wherein the first port, the second port and the front Each third port includes a full-bridge circuit containing four switching elements and an inductance component directly connected to the first winding, the second winding, or the third winding. The control circuit performs the step of calculating a control value using an approximate current waveform of the port current with respect to a target current. The approximate current waveform is derived from an approximate voltage waveform obtained by approximating the port voltage output from the full-bridge circuit with a continuous function. The approximate voltage waveform is an approximate function obtained by superimposing the fundamental wave and harmonics on the rectangular wave of the port voltage.

[0009] The second invention of this application is a triple active bridge converter of the first invention, wherein the approximate voltage waveform is an approximate function obtained by superimposing the fundamental wave and the third harmonic on the port voltage, which is a square wave.

[0010] The third invention of this application is a triple active bridge converter of the first invention, wherein the approximate voltage waveform is an approximate function obtained by superimposing the fundamental wave and the third harmonic on the port voltage, which is a square wave.

[0011] The fourth invention of this application is a triple active bridge converter according to any one of the first to third inventions, wherein step A) includes a) a step of calculating a Jacobian matrix, which is an approximate model at each operating point, for the approximate current waveform of the port current with respect to the target current; b) a step of calculating an inverse Jacobian matrix, which is the inverse of the Jacobian matrix; and c) a step of calculating a control value from the target current and the inverse Jacobian matrix.

[0012] The fifth invention of this application is a triple active bridge converter according to any one of the first to third inventions, wherein in step A), the control circuit calculates the control value by referring to a table that includes the correspondence between the target current, the detected port current, and the control value to be output, and the table is created from the results of deriving in advance under multiple conditions the following steps: a) calculate the Jacobian matrix, which is an approximate model at each operating point for the approximate current waveform of the port current relative to the target current; b) calculate the inverse Jacobian matrix, which is the inverse of the Jacobian matrix; and c) calculate the control value from the target current and the inverse Jacobian matrix.

[0013] The sixth invention of this application is a triple active bridge converter according to any one of the first to fifth inventions, wherein the inductance component is the leakage inductance between the actual element and the transformer, or a combination of the actual element and the transformer.

[0014] The seventh invention of this application is a multiactive bridge converter having three or more n ports, comprising: a transformer having n windings and a core that magnetically couples the n windings to each other; n ports connected to each of the n windings; and a control circuit for switching control of the switching elements of each of the ports, wherein each port includes a full bridge circuit including four switching elements and an inductance component directly connected to the winding, and the control circuit performs the step of A) calculating a control value using an approximate current waveform of the port current with respect to a target current, the approximate current waveform is derived from an approximate voltage waveform obtained by approximating the port voltage output from the full bridge circuit with a continuous function, and the approximate voltage waveform is an approximate function obtained by superimposing the fundamental wave and harmonics on the port voltage, which is a square wave. [Effects of the Invention]

[0015] According to the first to sixth inventions of this application, interference between two output ports can be suppressed more effectively compared to the case where a conventional approximation function is used to approximate the port voltage output from a full-bridge circuit using only the fundamental wave.

[0016] In particular, according to the second and third inventions, the interference suppression effect can be efficiently enhanced without excessively increasing the amount of computation.

[0017] According to the seventh invention of this application, interference between multiple output ports can be suppressed more effectively compared to the case where a conventional approximation function is used to approximate the port voltage output from a full-bridge circuit using only the fundamental wave. [Brief explanation of the drawing]

[0018] [Figure 1] This is a circuit diagram of the TAB converter according to the embodiment. [Figure 2] This block diagram schematically shows the signal flow in the control circuit and the modeled TAB converter. [Figure 3] This figure shows three approximate models of a square wave. [Figure 4] This is a flowchart showing the flow of the decoupling control method for the TAB converter according to the embodiment. [Figure 5] This diagram conceptually illustrates the operating characteristics of an integrator based on the integration of changes in the phase shift angle. [Figure 6] This figure shows the results of Simulation 1. [Figure 7] This figure shows the results of Simulation 2. [Figure 8] This figure shows the results of Simulation 3. [Figure 9] This figure shows the results of Experiment 1. [Figure 10] This figure shows the results of Experiment 2. [Figure 11] This figure shows the results of Experiment 3. [Modes for carrying out the invention]

[0019] The embodiments of the present invention will be described below with reference to the drawings.

[0020] <1. Circuit configuration of the TAB converter> Figure 1 is a circuit diagram of the TAB converter 1 according to this embodiment. The TAB converter 1 is a triple active bridge converter in which each of the three ports has a full bridge circuit.

[0021] The TAB converter 1 comprises a transformer T, a first port 10, a second port 20, a third port 30, and a control circuit 90. The first port 10 has a pair of first input / output terminals IO11 and IO12. The second port 20 has a pair of second input / output terminals IO21 and IO22. The third port 30 has a pair of third input / output terminals IO31 and IO32. A first DC power supply E1 is connected to the pair of first input / output terminals IO11 and IO12. A second DC power supply E2 is connected to the pair of second input / output terminals IO21 and IO22. A third DC power supply E3 is connected to the pair of third input / output terminals IO31 and IO32.

[0022] The TAB converter 1 transforms the power supply voltage of the first DC power supply E1, which is input from the first input / output terminals IO11 and IO12 of the first port 10, and outputs it from the second input / output terminals IO21 and IO22 and the third input / output terminals IO31 and IO32. In this case, the first port 10 becomes the input port, and the second port 20 and third port 30 become the output ports. Note that the TAB converter 1 does not necessarily transmit power in one direction; one or two of the first port 10, second port 20, and third port 30 may become input ports, and the remaining two or one may become output ports.

[0023] The transformer T comprises a first winding n1, a second winding n2, a third winding n3, and a core Tc. The first winding n1, the second winding n2, and the third winding n3 are magnetically coupled to each other via the core Tc. The first winding n1 is connected to the first port 10.

[0024] The first port 10 includes a first full-bridge circuit 100, a first DC power supply E1, a capacitor Co1, and a first inductor L1. The first DC power supply E1 and the first capacitor Co1 are connected in parallel between the two first input / output terminals IO11 and IO12. Alternatively, another voltage source, such as a voltage-controlled converter, may be used instead of the first DC power supply E1.

[0025] The first full-bridge circuit 100 has a first leg in which switching elements Q11 and Q12 are connected in series, and a second leg in which switching elements Q13 and Q14 are connected in series. The first and second legs are connected between two first input / output terminals IO11 and IO12.

[0026] Diodes D11, D12, D13, D14 and capacitors C11, C12, C13, C14 are connected in parallel to the switching elements Q11, Q12, Q13, Q14. Switching elements Q11 to Q14 are MOS-FETs. However, switching elements Q11 to Q14 may also be IGBTs or JFETs, etc. Diodes D11 to D14 may be solid elements or parasitic diodes. Capacitors C11 to C14 may be solid elements, parasitic capacitances, or a combination of parasitic capacitances and solid elements.

[0027] The first winding n1 of the transformer T is connected between the midpoint of the first leg and the midpoint of the second leg. As a result, the first winding n1 is connected to the input / output terminals IO11 and IO12 via the first full-bridge circuit 100. An inductor L1 is provided between the first winding n1 of the transformer T and the midpoint of the first leg. However, the inductor L1 may also be provided between the first winding n1 and the midpoint of the second leg. Furthermore, the inductor L1 may be divided and arranged between the first winding n1 and the midpoint of the first leg, and between the first winding n1 and the midpoint of the second leg. The inductor L1 may be a physical element, the leakage inductance of the transformer T, or a combination of a physical element and the leakage inductance.

[0028] The second port 20 includes a first full-bridge circuit 200, a second DC power supply E2, a capacitor Co2, and a second inductor L2. The second DC power supply E2 and the second capacitor Co2 are connected in parallel between the two first input / output terminals IO21 and IO22. Alternatively, another voltage source, such as a voltage-controlled converter, may be used instead of the second DC power supply E2.

[0029] The second full-bridge circuit 200 has a third leg in which switching elements Q21 and Q22 are connected in series, and a fourth leg in which switching elements Q23 and Q24 are connected in series. The third and fourth legs are connected between two second input / output terminals IO21 and IO22.

[0030] Switching elements Q21, Q22, Q23, and Q24 are connected in parallel to diodes D21, D22, D23, and D24, and capacitors C21, C22, C23, and C24. Switching elements Q21 to Q24 are MOS-FETs. However, switching elements Q21 to Q24 may also be IGBTs or JFETs, etc. Diodes D21 to D24 may be solid elements or parasitic diodes. Capacitors C21 to C24 may be solid elements, parasitic capacitances, or a combination of parasitic capacitances and solid elements.

[0031] The second winding n2 of transformer T is connected between the midpoint of the third leg and the midpoint of the fourth leg. As a result, the second winding n2 is connected to input / output terminals IO21 and IO22 via the second full-bridge circuit 200. A second inductor L2 is provided between the second winding n2 of transformer T and the midpoint of the third leg. However, the second inductor L2 may be provided between the second winding n2 and the midpoint of the fourth leg. Alternatively, the inductor L2 may be divided and arranged between the second winding n2 and the midpoint of the third leg, and between the second winding n2 and the midpoint of the fourth leg. The second inductor L2 may be a real element, the leakage inductance of transformer T, or a combination of a real element and the leakage inductance.

[0032] The third port 30 includes a third full-bridge circuit 300, a third DC power supply E3, a capacitor Co3, and a third inductor L3. The third DC power supply E3 and the third capacitor Co3 are connected in parallel between the two first input / output terminals IO31 and IO32. Alternatively, another voltage source, such as a voltage-controlled converter, may be used instead of the third DC power supply E3.

[0033] The third full-bridge circuit 300 has a fifth leg in which switching elements Q31 and Q32 are connected in series, and a sixth leg in which switching elements Q33 and Q34 are connected in series. The fifth and sixth legs are connected between two third input / output terminals IO31 and IO32.

[0034] Switching elements Q31, Q32, Q33, and Q34 are connected in parallel to diodes D31, D32, D33, and D34, and capacitors C31, C32, C33, and C34. Switching elements Q31 to Q34 are MOS-FETs. However, switching elements Q31 to Q34 may also be IGBTs or JFETs, etc. Diodes D31 to D34 may be solid elements or parasitic diodes. Capacitors C31 to C34 may be solid elements, parasitic capacitances, or a combination of parasitic capacitances and solid elements.

[0035] The third winding n3 of transformer T is connected between the midpoint of the fifth leg and the midpoint of the sixth leg. As a result, the third winding n3 is connected to input / output terminals IO31 and IO32 via the third full-bridge circuit 300. A third inductor L3 is provided between the third winding n3 of transformer T and the midpoint of the fifth leg. However, the third inductor L3 may be provided between the third winding n3 and the midpoint of the sixth leg. Alternatively, the inductor L3 may be divided and arranged between the third winding n3 and the midpoint of the fifth leg, and between the third winding n3 and the midpoint of the sixth leg. The third inductor L3 may be a real element, the leakage inductance of transformer T, or a combination of a real element and the leakage inductance.

[0036] The gate terminals of switching elements Q11-Q14, Q21-Q24, and Q31-Q34 are wired to receive signals output from the control circuit 90. The control circuit 90 controls the switching of switching elements Q11-Q14 of the first port 10, Q21-Q24 of the second port 20, and Q31-Q34 of the third port 30 so that the output power of the TAB converter 1 reaches the set target power.

[0037] In such a TAB converter 1, the control circuit 90 performs switching control on the first full-bridge circuit 100 of the first port 10, the second full-bridge circuit 200 of the second port 20, and the third full-bridge circuit 300 of the third port 30. As a result, rectangular wave port voltages Vx, Vy, and Vz are generated across the inductors L1, L2, and L3 of each port 10, 20, and 30. The TAB converter 1 performs power transmission between ports by controlling the phase difference of these port voltages Vx, Vy, and Vz.

[0038] <2. Interference in TAB converters and conventional decoupling> Before describing the control method for the TAB converter that performs decoupling in this embodiment, we will first explain the interference between the second and third ports in the TAB converter and conventional decoupling methods.

[0039] The TAB converter 1 has three ports, the first port 10, the second port 20, and the third port 30, which are magnetically connected via a transformer T, while being electrically isolated. Furthermore, the input and output are variable within the range where the sum of the powers of the first port 10, the second port 20, and the third port 30 is zero. Power can be transmitted between ports through the phase difference of the first full-bridge circuit 100, the second full-bridge circuit 200, and the third full-bridge circuit 300.

[0040] If the phase difference between the second port 20 and the first port 10 is controlled by considering only the output current of the second port 20, the output current of not only the second port 20 but also the third port 30 will fluctuate. This causes interference between the second port 20 and the third port 30. Therefore, it is necessary to "de-interfere" the ports. Specifically, "de-interfere" refers to controlling the manipulated amount of the port where interference occurs in order to cancel out the interference.

[0041] In the aforementioned Non-Patent Document 1, the approximate current waveforms of the port currents flowing through each port 10, 20, and 30 are derived from approximate voltage waveforms obtained by approximating the port voltages Vx, Vy, and Vz of each port 10, 20, and 30, which are originally rectangular waves, with the fundamental wave. Then, the derived approximate current waveforms are decoupled using a linear approximation model at each operating point. When the port voltages Vx, Vy, and Vz are approximated with the fundamental wave, the approximate current waveforms of the current I2 flowing through the second port 20 and the current I3 flowing through the third port 30 are expressed as follows, using the phase shift angle φ2 of the second port 20 with respect to the first port 10 and the phase shift angle φ3 of the third port 30 with respect to the first port 10.

number

number

Count

[0043] , 11 , , ,

[0044] , , , , -1 , K 12 , K 13 , K 14 is as follows.

Count

[0045] <3. Control method of the TAB converter in this embodiment> Next, the control method for the TAB converter 1 in this embodiment will be described with reference to Figure 2. Figure 2 is a schematic block diagram showing the signal flow in the control circuit 90 and the modeled TAB converter 1. The left side of Figure 2 shows the signal flow in the control circuit 90 of this embodiment. The right side of Figure 2 is Model M, which models the signal movement in the TAB converter 1.

[0046] This control circuit 90 calculates the phase shift angle φ2 of the second port 20 relative to the first port 10 and the phase shift angle φ3 of the third port 30 relative to the first port 10 using the inverse Jacobian matrix calculated using Model M, and performs decoupling control on the TAB converter 1 by inputting control signals based on these phase shift angles φ2 and φ3 to the TAB converter 1.

[0047] The control circuit 90 has a port current I for the second port 20, as shown on the left end of Figure 2. tab2 Output current command value I tab2 * And the port current I of the third port 30 tab3 Output current command value I tab3 * The following is input. The control circuit 90 calculates the phase shift angle φ2 of the second port 20 relative to the first port 10, and the phase shift angle φ3 of the third port 30 relative to the first port 10.

[0048] As shown in the center of Figure 2, the control circuit 90 inputs control signals to the model M based on the calculated phase shift angles φ2 and φ3. When actually controlling the TAB converter 1, the control circuit 90 inputs control signals to the TAB converter 1 based on the calculated phase shift angles φ2 and φ3. Specifically, the control circuit 90 outputs drive signals to each of the switching elements Q11~Q14, Q21~Q24, and Q31~Q34 included in the first full-bridge circuit 100, the second full-bridge circuit 200, and the third full-bridge circuit 300, respectively, to control their switching. As a result, the second port 20 and the third port 30 of the TAB converter 1 each receive a port current I tab2 ,I tab3 It plays.

[0049] Here, we will explain the design method for Model M of the TAB converter 1 shown in Figure 2. Model M is the port current I that flows through the second port 20 and third port 30 of the TAB converter 1 when switching signals corresponding to the phase shift angles φ2 and φ3 are input to the TAB converter 1. tab2 ,I tab3 This is calculated using an approximate model. Model M is a model shown using the Jacobian matrix G described later. In this model M, the phase shift angle φ2 of the second port 20 relative to the first port 10 and the phase shift angle φ3 of the third port 30 relative to the first port 10 are taken as inputs, and the port current I of the second port 20 and the third port 30 is calculated. tab2 ,I tab3 Outputs.

[0050] As shown in Figure 2, Model M has differentiators 81a, 81b, a Jacobian matrix operation unit 82, and integrators 83a, 83b. Model M takes the phase shift angles φ2 and φ3 of the switching signal input to the TAB converter 1 as input.

[0051] In the TAB converter 1, the change amounts Δφ2, Δφ3 of the phase shift angles φ2, φ3 of the input switching signals and the port current I flowing through the second port 20 and the third port 30 are used. tab2 ,I tab3Change ΔI tab2 ,ΔI tab3 The relationship between them is as follows:

number

[0052] Model M utilizes this relationship. In Model M, the input phase shift angles φ2 and φ3 are differentiated in differentiators 81a and 81b, respectively, to calculate the changes in phase shift angles φ2 and φ3, Δφ2 and Δφ3. Then, in the Jacobian matrix calculation unit 82, the Jacobian matrix G is integrated into these changes to obtain the port current I as shown in the above equation. tab2 ,I tab3 Change ΔI tab2 ,ΔI tab3 We obtain the following. In this case, the Jacobian matrix G is expressed by the following equation.

number

[0053] Next, the change ΔI tab2 ,ΔI tab3 The values ​​are integrated in integrators 83a and 83b to obtain the port current I tab2 ,I tab3 You can obtain this.

[0054] In this model M, the Jacobian matrix G is the port current I tab2 ,I tab3 The current waveform is calculated using an approximate current waveform, which is an approximate model of the target current waveform. In this case, the smaller the error between the target current waveform and the approximate current waveform, the more accurately the decoupling can be performed in the control circuit 90. On the other hand, the more complex the approximate model, the greater the computational load in the control circuit 90, which puts a strain on resources. For this reason, it is preferable to use an approximate model that has a good balance between the magnitude of the error and the amount of computation.

[0055] Figure 3 shows three types of approximation models for a square wave. Figure 3 shows a square wave, a model approximating the square wave using only the fundamental wave, a model approximating the square wave by superimposing the fundamental wave and the third harmonic, and a model approximating the square wave by superimposing the fundamental wave, the third harmonic, and the fifth harmonic. As shown in Figure 3, the approximation model using only the fundamental wave has a large error compared to the square wave, whereas the approximation models with superimposed harmonics show a decrease in error as the number of superimposed harmonics increases. On the other hand, as the number of superimposed harmonics increases, the processing load for calculating the control signal in the control circuit 90 increases.

[0056] Therefore, in the control circuit 90 of this embodiment, a continuous function with the fundamental wave and harmonics superimposed is used as an approximate voltage waveform, which is an approximate model of the port voltages Vx, Vy, Vz of each port 10, 20, 30, which are rectangular waves. Then, from the approximate voltage waveform which is a continuous function, the port current I tab2 ,I tab3 An approximate current waveform is derived. This approximate current waveform is used to perform decoupling control to suppress interference between ports. As a result, the error between the actual port voltages Vx, Vy, Vz and the approximate voltage waveform of the port voltages Vx, Vy, Vz is reduced compared to when decoupling control is performed using an approximation function expressed only by the fundamental wave. Therefore, the port current I derived from the approximate voltage waveform is tab2 ,I tab3 The error due to approximation of the approximate current waveform is also reduced. As a result, interference between ports can be further suppressed.

[0057] Specifically, for the rectangular wave port voltages Vx, Vy, and Vz, either a function that superimposes the fundamental wave and the third harmonic, or a function that superimposes the fundamental wave, the third harmonic, and the fifth harmonic, is used as the approximate voltage waveform. Below, we will explain the case where a function that approximates the voltage waveform of each port voltage Vx, Vy, and Vz by superimposing the fundamental wave and the third harmonic is used.

[0058] Here, the port current I is obtained when the port voltages Vx, Vy, Vz are approximated by a function obtained by superimposing the fundamental wave (1st order) and odd-order harmonics. tab2 ,Itab3 The approximate current waveform is expressed by the following equation.

number

number

number

number

number

[0059] Note that n represents the number of superimposed sine waves in the approximate voltage waveform. In other words, when n=1, the model is approximated by a single sine wave, i.e., only the fundamental wave. When n=2, the model is approximated by two sine waves, i.e., the fundamental wave and the third harmonic, superimposed. And when n=3, the model is approximated by three sine waves, i.e., the fundamental wave, the third harmonic, and the fifth harmonic, superimposed.

[0060] When n=2, i.e., in the case of an approximate model of the fundamental wave and the third harmonic, the port current I shown in Equations 12 and 13 is tab2 ,I tab3 The approximate current waveform is as follows.

number

number

[0061] Port current I in the approximate model tab2 ,I tab3 Taking the partial derivatives of this with respect to φ2 and φ3, respectively, yields the following:

number

number

number

number

[0062] From Equation 11, the components of the Jacobian matrix G are defined as follows:

number

number

number

number

number

[0063] In the model M shown in Figure 2, the Jacobian matrix operation unit 82 includes matrix component multiplication units 82a, 82b, 82c, 82d, adders 82e, 82f, and coefficient multiplication units 82g, 82h.

[0064] The matrix component multiplication unit 82a multiplies the change in the phase shift angle φ2 Δφ2 by the component G of the Jacobian matrix G. 11 Multiply by Δφ2·G 11 The matrix component multiplication section 82b calculates the component G of the Jacobian matrix G multiplied by the change in phase shift angle φ3 Δφ3. 12 Multiply by Δφ3·G 12The matrix component multiplication section 82c multiplies the change in the phase shift angle φ2 Δφ2 by the components G of the Jacobian matrix G. 21 Multiply by Δφ2·G 21 The matrix component multiplication section 82d calculates the component G of the Jacobian matrix G multiplied by the change in the phase shift angle φ3 Δφ3. 22 Multiply by Δφ3·G 22 Calculate.

[0065] Δφ2·G calculated in the matrix component multiplication section 82a, 84b 11 and Δφ3·G 12 These are added in the adder 82e, and then multiplied by a coefficient in the multiplier 82g, resulting in 8 / π 2 ω is multiplied. Meanwhile, Δφ2·G calculated in the matrix component multiplication section 82c, 84d 21 and Δφ3·G 22 These are added in adder 82f, and then multiplied by coefficients 8 / π in coefficient multiplication unit 82h. 2 ω is multiplied. As a result, the port current I shown in the following equation tab2 ,I tab3 Change ΔI tab2 ,ΔI tab3 To obtain.

number

[0066] Next, we will explain the method of decoupling control in the control circuit 90 using such a model M.

[0067] Figure 4 is a flowchart showing the flow of the control method for the TAB converter 1, taking decoupling into consideration, in the control circuit 90 of this embodiment. In order to perform decoupling control in the control circuit 90 shown in Figure 4, an approximate function is prepared in which the fundamental wave and the third harmonic are superimposed as the approximate voltage waveforms of each port voltage Vx, Vy, and Vz.

[0068] When controlling the TAB converter 1, as shown in Figure 4, the control circuit 90 repeatedly performs steps S1 to S4 for each sampling period. First, the Jacobian matrix G, which is an approximate model using differentiation, is calculated (step S1). This gives the changes in the phase shift angles φ2 and φ3 in the approximate model, Δφ2 and Δφ3, and the port current I output from the model M. tab2 ,I tab3 Change ΔI tab2 ,ΔI tab3 The relationship is calculated.

[0069] Here, as shown in equation 28, the port current I of the second port 20 tab2 This is interfered with by the change in the phase shift angle φ3 of the third port 30, and the port current I of the third port 30 tab3 This is interfered with by the change in the phase shift angle φ2 of the second port 20. To cancel this interference, the inverse Jacobian matrix H is calculated in the next step S2.

[0070] After calculating the Jacobian matrix G, the control circuit 90 calculates the inverse Jacobian matrix H (step S2). The inverse Jacobian matrix H is the inverse of the Jacobian matrix G, and when multiplied by the Jacobian matrix G, it becomes the identity matrix. In other words, the inverse Jacobian matrix H is the inverse model of the Jacobian matrix G, which is an approximate model obtained by differentiation. The Jacobian matrix G and the inverse Jacobian matrix H have the following relationship.

number

[0071] Each component H of the inverse Jacobian matrix H 11 ,H 12 ,H 21 ,H 22 It can be calculated as follows:

number

[0072] In the actual operation of the control circuit 90, steps S1 and S2 are integrated, and each component G of the Jacobian matrix G is processed. 11 ,G12 , G 21 , G 22 Without calculating, the relationship between numbers 23 to 27 and number 30 is used to directly calculate each component of the inverse Jacobian matrix H 11 , H 12 , H 21 , H 22 .

[0073] When the inverse Jacobian matrix H is calculated, the control circuit 90 uses the inverse Jacobian matrix to calculate the output current command values I tab2 of the port current I tab2 * of the second port 20 and the output current command values I tab3 of the port current I tab3 * of the third port 30, and based on these, calculates the phase shift angles φ2 and φ3 which are control values (step S3).

[0074] Here, while referring to FIG. 2, the calculation procedure of the control value in the control circuit 90 will be described. In FIG. 2, as control blocks corresponding to the operation of the control circuit 90, regulators 91a, 91b, an inverse Jacobian matrix calculation unit 92, and integrators 93a, 93b are shown.

[0075] In step S3, first, the target change amounts ΔI tab2 , ΔI tab3 of the port currents I tab2 , ΔI tab3 of the second port 20 and the third port 30 are calculated (step S31). Specifically, as shown in FIG. 2, the control circuit 90 inputs the output current command value I tab2 of the port current I tab2 * of the second port 20 and the current detection value of the actual current I tab2 to the regulator 91a. The regulator 91a uses the output current command value I tab2 * and the current detection value of the actual current I tab2 to calculate the target change amount ΔI tab2 of the port current I tab2 .

[0076] Furthermore, the control circuit 90 controls the port current I of the third port 30. tab3 Output current command value I tab3 * and actual current I tab3 The detected current value is input to regulator 91b. Regulator 91b outputs current command value I tab3 * and actual current I tab3 Using the current detection value, the port current I tab3 Target change ΔI tab3 Calculate.

[0077] Next, the control circuit 90 calculates the change amounts Δφ2 and Δφ3 of the phase shift angles φ2 and φ3 (step S32). In step S32, the control circuit 90 calculates the target change amount ΔI in the inverse Jacobian matrix calculation unit 92. tab2 ,ΔI tab3 This is multiplied by the inverse Jacobian matrix H. This calculates the changes in phase shift angles φ2 and φ3, Δφ2 and Δφ3. That is, in step S32, the calculated target change ΔI tab2 ,ΔI tab3 Using the formulas for the inverse Jacobian matrix H, which is the inverse model of the Jacobian matrix G, we calculate the changes in phase shift angles φ2 and φ3, Δφ2 and Δφ3. The changes in phase shift angles φ2 and φ3, Δφ2 and Δφ3, are expressed by the following equations.

number

[0078] Specifically, in the control block diagram of the control circuit 90 shown in Figure 2, the inverse Jacobian matrix operation unit 92 includes matrix component multiplication units 92a, 92b, 92c, 92d and adders 92e, 92f.

[0079] The matrix component multiplication unit 92a calculates the target change amount ΔI tab2 The component H of the inverse Jacobian matrix H 11 Multiply by ΔI tab2 ·H 11 The matrix component multiplication unit 92b calculates the calculated target change ΔI. tab3 The component H of the inverse Jacobian matrix H 12 Multiply by ΔItab3 ·H 12 The matrix component multiplication unit 92c calculates the calculated target change ΔI. tab2 The component H of the inverse Jacobian matrix H 21 Multiply by ΔI tab2 ·H 21 The matrix component multiplication unit 92d calculates the target change ΔI. tab3 The component H of the inverse Jacobian matrix H 22 Multiply by ΔI tab3 ·H 22 Calculate.

[0080] ΔI calculated in the matrix component multiplication section 92a, 92b tab2 ·H 11 and ΔI tab3 ·H 12 This is added by adder 92e. Meanwhile, ΔI calculated in matrix component multiplication units 92c and 92d tab2 ·H 21 and ΔI tab3 ·H 22 These are added by adder 92f. This allows us to calculate the changes in phase shift angles φ2 and φ3, Δφ2 and Δφ3, as shown in the following equation.

number

[0081] Then, the control circuit 90 inputs the change amounts Δφ2 and Δφ3 of the phase shift angles φ2 and φ3 output from the inverse Jacobian matrix calculation unit 92 to the integrators 93a and 93b, respectively, and calculates the control values ​​of the phase shift angles φ2 and φ3 (step S33). The integrators 93a and 93b calculate the phase shift angles φ2 and φ3 by integrating the change amounts as follows. Note that in the following, φ in the period is 2(n) ,φ 3(n) This shows the values ​​of φ2 and φ3 in the nth operation, and Δφ 2(n) ,Δφ 3(n) This shows the values ​​of Δφ2 and Δφ3 in the nth operation.

number

number

[0082] In other words, in step S33, all the change amounts Δφ2 and Δφ3 calculated in previous steps S32 during the period from the start of the calculation to the present, and the change amounts Δφ2 and Δφ3 calculated in the current (new) step S32 are added together to calculate the phase shift angles φ2 and φ3.

[0083] In addition, in integrators 93a and 93b, new phase shift angles φ2 and φ3 may be calculated by adding the phase shift angles φ2 and φ3 calculated in the previous step S33 and the change amounts Δφ2 and Δφ3 calculated in the current (new) step S32. In that case, the values ​​of φ2 and φ3 in the nth calculation are indicated by φ 2(n) ,φ 3(n) It is calculated as follows:

number

number

[0084] Here, Figure 5 conceptually shows the operating characteristics obtained by integrating the changes in phase shift angles φ2 and φ3 Δφ2 and Δφ3 in integrators 93a and 93b. In Figure 5, the horizontal axis represents the phase shift angles φ2 and φ3, and the vertical axis represents the output current, i.e., the port current I. tab2 ,I tab3 This is shown. In Figure 5, the current trajectory of the operating characteristics of TAB converter 1 is shown as a solid line, and the current trajectory when the above integration is performed is shown as a dashed line. Also, in Figure 5, the change amount Δφ at n=4 is shown. (4) The current trajectory when integration is not performed is shown by the thick dotted line, and its extension is shown by the thin dotted line.

[0085] Δφ (n) If, after the calculation, integration is not performed using integrators 93a and 93b, then Δφ (n) And, ΔI tabThe relationship between the two points determines the operating point through a linear approximation formula that passes through the origin. Therefore, in the case of n=4, the point on the linear approximation line shown by the thick and thin dashed lines is the operating point (the "operating point without integration" in Figure 5).

[0086] In contrast, in this embodiment, Δφ (n) By performing integration in integrators 93a and 93b each time an operation is performed, the intercept of the approximation line in the model using the Jacobian matrix G, which is a linear approximation model, can be moved from the origin to the integrated operating point (the "integrated operating point" in Figure 5), as shown by the dashed line in Figure 5. This reduces the error with the operating characteristics, resulting in an operating point equivalent to the target current trajectory and obtaining a trajectory with the same phase. As a result, stable power characteristics can be obtained even under heavy load.

[0087] After the phase shift angles φ2 and φ3 are calculated in step S3, the control circuit 90 outputs drive signals to each of the switching elements Q11~Q14, Q21~Q24, and Q31~Q34 included in the first full-bridge circuit 100, second full-bridge circuit 200, and third full-bridge circuit 300 of the TAB converter 1 based on the phase shift angles φ2 and φ3, thereby controlling their switching (step S4).

[0088] By performing this deinterfering control, the port current I between the second port 20 and the third port 30 is tab2 ,I tab3 This method can suppress interference more effectively than conventional methods.

[0089] In the description of the above embodiment, the control circuit 90 is described as "calculating a matrix," but the control circuit 90 may calculate it in matrix form, or it may calculate each component of the matrix. Similarly, the matrix calculation formulas shown in each mathematical formula may be calculated in matrix form, or the components of the matrix to be calculated may be calculated from the components of the matrix before the calculation.

[0090] <4. Simulation> <4-1. Simulation 1> To confirm the effect of the decoupling control in the above embodiment, simulations were performed under the following conditions: when the inverse Jacobian matrix H is not multiplied in the control circuit 90 (hereinafter referred to as "with interference") and when the inverse Jacobian matrix H is multiplied (hereinafter referred to as "decoupling"). [Common conditions] Target current value for port 20: Step change from 0.5A to 3.0A. The target current value for port 30 remains at 0.5A. [Decoherent control] Approximate model: Fundamental wave + 3rd harmonic Whether or not the change in phase shift angle is integrated: Yes

[0091] Figure 6 shows the results of Simulation 1. In this Simulation 1, the port current I of the second port 20 is tab2 If this is changed, the port current I of the third port 30 tab3 This allows us to observe what changes occur depending on whether or not decoupling control is present. The upper part of Figure 6 shows the port current I of the second port 20. tab2 Regarding this, the lower part of Figure 6 shows the port current I of the third port 30. tab3 For each case, the current command value, the case with interference, and the case with de-interference are shown.

[0092] As shown in Figure 6, the port current I of the second port 20 tab2 When current fluctuations occur, if interference is present, the port current I in the third port 30 tab3 The current temporarily decreased by approximately 0.45A. In contrast, when deinterference was performed, the port current I in the third port 30 decreased. tab3 The current temporarily increased by approximately 0.08A before decreasing by approximately 0.03A. This indicates that when de-interference is performed, the current fluctuation in the third port 30 caused by interference from current fluctuations in the second port 20 is significantly smaller compared to when interference is present.

[0093] <4-2. Simulation 2> Next, to confirm the effect of decoupling due to differences in approximation models, simulations were performed for three types of approximation models under the following conditions. [conditions] Approximation models: 3 types • Fundamental wave only ·Fundamental wave + 3rd harmonic ·Fundamental wave + 3rd harmonic + 5th harmonic Target current value for port 20: Step change from 0.5A to 3.5A. The target current value for port 30 remains at 0.5A. Whether or not the change in phase shift angle is integrated: Yes

[0094] Figure 7 shows the results of Simulation 2. In Figure 7, the left column shows the results when the approximation model is the fundamental wave only, the middle column shows the results when the approximation model is the superposition of the fundamental wave and the third harmonic, and the right column shows the results when the approximation model is the superposition of the fundamental wave, the third harmonic, and the fifth harmonic. Also, in each column of Figure 7, the top row shows the port current I of the second port 20. tab2 The lower section shows the port current I of the third port 30. tab3 This shows that in Simulation 2, the port current I of the third port 30 is shown. tab3 The fluctuations are as follows: [Simulation Results] Fundamental frequency only: 0.025A Fundamental wave + 3rd harmonic: 0.021A Fundamental wave + 3rd harmonic + 5th harmonic: 0.017A

[0095] The results of Simulation 2 show that as the order of harmonic superposition increases, the influence of interference on the third port 30 from the current value change of the second port 20 decreases. In other words, by using the inverse Jacobian matrix H of the harmonic superposition model in the control circuit 90, interference between the two output ports can be further suppressed in the TAB converter 1.

[0096] <4-3. Simulation 3> Furthermore, the port current I depends on whether or not the changes in phase shift angles φ2 and φ3 in integrators 93a and 93b are integrated. tab2 ,I tab3 To compare the operational stability of the system, a simulation was performed under the following conditions. [conditions] Approximate model: Fundamental wave + 3rd harmonic Target current value for port 20: Step change from 0.5A to 5.0A (equivalent to heavy load) The target current value for port 30 remains at 0.5A. Integration of the change in phase shift angle: 2 types - None / Yes

[0097] Figure 8 shows the results of Simulation 3. In Figure 8, the left column shows the results when the change in the phase shift angle is not integrated ("No Integration" in Figure 8), and the right column shows the results when the change in the phase shift angle is integrated ("Integration" in Figure 8). Also, in each column of Figure 8, the top row shows the port current I of the second port 20. tab2 The lower section shows the port current I of the third port 30. tab3 This indicates that.

[0098] As shown in Figure 8, without integration, the control diverged in the heavy load region, and it was not possible to perform operation in accordance with the command value. In contrast, with integration, stable operation was possible. In other words, by integrating the changes in phase shift angles φ2 and φ3, Δφ2 and Δφ3, in the control circuit 90, more stable decoupling control can be performed in the TAB converter 1.

[0099] <5. Experiment> <5-1. Experiment 1> To confirm the effect of the decoupling control in the above embodiment, experiments were conducted using the TAB converter 1 under the following conditions: when the inverse Jacobian matrix H was not integrated in the control circuit 90 (hereinafter referred to as "with interference") and when the inverse Jacobian matrix H was integrated (hereinafter referred to as "decoupling"). [Common conditions] Target current value for port 20: Step change from 1.5A to 4.5A. The target current value for port 30 remains at 1.5A. [Decoherent control] Approximate model: Fundamental wave + 3rd harmonic Whether or not the change in phase shift angle is integrated: Yes

[0100] Figure 9 shows the results of Experiment 1. In Experiment 1, the port current I of the second port 20 was tab2 If this is changed, the port current I of the third port 30 tab3 This allows us to observe what changes occur depending on whether or not decoupling control is present. The upper part of Figure 9 shows the port current I of the second port 20 under the condition of interference. tab2 and the port current I of the third port 30 tab3 Regarding this, the lower part of Figure 9 shows the port current I of the second port 20 under decoupling conditions. tab2 and the port current I of the third port 30 tab3 The current waveform averaged over the sampling period is shown.

[0101] As shown in Figure 9, the port current I of the second port 20 tab2 When current fluctuations occur, if interference is present, the port current I in the third port 30 tab3 The current temporarily dropped by approximately 1.32A. In contrast, when deinterference was performed, the port current I at the third port 30 decreased. tab3 The temporary fluctuation amount was 58.0mA. From this, it can be seen that when de-interference is performed, the value of the current fluctuation in the third port 30 caused by interference from the current fluctuation in the second port 20 is significantly smaller compared to when interference is present.

[0102] <5-2. Experiment 2> Next, to confirm the effect of decoupling due to differences in approximation models, experiments were conducted for three types of approximation models under the following conditions. Other conditions were the same as those for the decoupling control in Experiment 1. [conditions] Approximation models: 3 types • Fundamental wave only ·Fundamental wave + 3rd harmonic ·Fundamental wave + 3rd harmonic + 5th harmonic Target current value for port 20: Step change from 1.5A to 4.5A. The target current value for port 30 remains at 1.5A. Whether or not the change in phase shift angle is integrated: Yes

[0103] Figure 10 shows the results of Experiment 2. In Figure 10, the top panel shows the results when the approximation model is the fundamental wave only, the middle panel shows the results when the approximation model is the superposition of the fundamental wave and the third harmonic, and the bottom panel shows the results when the approximation model is the superposition of the fundamental wave, the third harmonic, and the fifth harmonic. In Experiment 2, the port current I of the third port 30 tab3 The fluctuations are as follows: [Experimental Results] Fundamental frequency only: 0.27A Fundamental wave + 3rd harmonic: 0.25A Fundamental wave + 3rd harmonic + 5th harmonic: 0.21A

[0104] The results of Experiment 2 show that as the order of harmonic superposition increases, the influence of interference on the third port 30 from the current value change of the second port 20 decreases. In other words, by using the inverse Jacobian matrix H of the harmonic superposition model in the control circuit 90, interference between the two output ports can be further suppressed in the TAB converter 1.

[0105] <5-3. Experiment 3> Next, the port current I depends on whether or not the changes in phase shift angles φ2 and φ3 in integrators 93a and 93b are integrated. tab2 ,I tab3 To compare the operational stability, the experiment was conducted under the following conditions. Other conditions were the same as those for the decoupling control in Experiment 1. [conditions] Approximate model: Fundamental wave + 3rd harmonic Target current values ​​for port 20: 2 types Step change from 0.50A to 3.50A Step change from 0.50A to 5.17A Maintain the target current value for port 30 (port 3): 0.50A. Integration of the change in phase shift angle: 2 types - None / Yes

[0106] Figure 11 shows the results of Experiment 3. In Figure 11, the left column shows the results when the change in the phase shift angle is not integrated ("No Integration" in Figure 11), and the right column shows the results when the change in the phase shift angle is integrated ("Integration" in Figure 11). In addition, in each column of Figure 11, the upper row shows the results under relatively light load conditions, and the lower row shows the results under relatively heavy load conditions. Note that the relatively light load condition refers to the port current I of the second port 20. tab2 This is the case where the current is stepped from 0.50A to 3.50A (load factor 54%). Also, a relatively heavy load is the port current I of the second port 20. tab2 This is the case where the current is changed in steps from 0.50A to 5.17A (80% load factor).

[0107] As shown in Figure 11, without integration, even under relatively light load conditions, the port current I of both the second port 20 and the third port 30 is higher compared to when integration is enabled. tab2 ,I tab3 The vibrations were increasing. Furthermore, under relatively heavy loads, without integration, the control diverged, and it was not possible to perform operations that followed the command value. In contrast, with integration, stable operation was possible even under heavy loads. From this, it can be said that by integrating the changes in phase shift angles φ2 and φ3, Δφ2 and Δφ3, in the control circuit 90, more stable decoupling control can be performed in the TAB converter 1.

[0108] <6. Variation> Although one embodiment of the present invention has been described above, the present invention is not limited to the above-described embodiment.

[0109] The above embodiment describes decoupling control of a triple active bridge converter having three ports, but the present invention is not limited thereto. The decoupling control method using the inverse Jacobian matrix of the present invention can also be applied to a multi-active bridge converter having four or more ports with full bridge circuits.

[0110] In the above embodiment, the Jacobian matrix G and the inverse Jacobian matrix H were calculated using an approximate model including harmonics, the changes in phase shift angles φ2 and φ3 Δφ2 and Δφ3 were calculated from the target current and the inverse Jacobian matrix, and decoupling control was performed by integrating the changes in phase shift angles φ2 and φ3 Δφ2 and Δφ3. However, the present invention is not limited thereto. Decoupling control may be performed using an approximate model including harmonics without integrating the changes in phase shift angles, or decoupling control may be performed using an approximate model of only fundamental waves while integrating the changes in phase shift angles.

[0111] Furthermore, instead of steps S1, S2, S31, and S32 of the above embodiment, the target current (port current I of the second port 20 and the third port 30) is used. tab2 ,I tab3 Output current command value I tab2 * ,I tab3 * ) and the detected port current (actual current I tab2 ,I tab2 ) and, referring to a pre-prepared table, determine the port current I of the second port 20 and the third port 30. tab2 ,I tab2 Target change ΔI tab2 ,ΔI tab3 The calculation may be performed in this way. In this case, the table is created using the results obtained in advance from the steps corresponding to steps S1, S2, S31, and S32 of the above embodiment, under multiple conditions.

[0112] Furthermore, in the above embodiment, the first port 10 was described as the input side, and the second port 20 and third port 30 were described as the output sides. However, the TAB converter 1 may use any of the three ports as the input port. In that case, the remaining port becomes the output port.

[0113] Furthermore, the elements that appear in the above embodiments or modifications may be combined as appropriate, to the extent that no contradictions arise. [Explanation of symbols]

[0114] 1 TAB Converter 10. Port 1 20 Port 2 30 Third Port 90 Control circuits 91a, 91b regulators 92 Inverse Jacobian matrix calculation section 93a,93b Integrator 100 First Full Bridge Circuit 200 Second Full Bridge Circuit 300 Third Full Bridge Circuit G Jacobian procession H Inverse Jacobian Matrix L1 First Inductor L2 Second Inductor L3 Third Inductor M Model T Transformer Tc Core n1 First winding n2 Second winding n3 Third winding ΔItab2, ΔItab3: Change in target current φ2, φ3 Phase shift angle Change in phase shift angle Δφ2, Δφ3

Claims

1. A triple active bridge converter having three ports, A transformer having a first winding, a second winding, a third winding, and a core that magnetically couples the first winding, the second winding, and the third winding together, The first port connected to the first winding, The second port connected to the second winding, The third port connected to the third winding, A control circuit for switching control the switching elements of the first port, the second port, and the third port, Equipped with, The first port, the second port, and the third port are, A full bridge circuit including four switching elements, The inductance component directly connected to the first winding, the second winding, or the third winding, Includes, The aforementioned control circuit is A) A process of calculating the control value using the approximate current waveform of the port current relative to the target current. Execute, The aforementioned approximate current waveform is derived from an approximate voltage waveform obtained by approximating the port voltage output from the full-bridge circuit with a continuous function. The aforementioned approximate voltage waveform is an approximate function obtained by superimposing the fundamental wave and harmonics onto the port voltage, which is a square wave, in a triple active bridge converter.

2. A triple active bridge converter according to claim 1, The aforementioned approximate voltage waveform is an approximate function obtained by superimposing the fundamental wave and third harmonic on the port voltage, which is a square wave, in a triple active bridge converter.

3. A triple active bridge converter according to claim 1, The aforementioned approximate voltage waveform is an approximation function obtained by superimposing the fundamental wave, third harmonic, and fifth harmonic on the port voltage, which is a square wave, in a triple active bridge converter.

4. A triple active bridge converter according to any one of claims 1 to 3, The aforementioned step A) is, a) A step of calculating the Jacobian matrix, which is an approximate model at each operating point, for the approximate current waveform of the port current with respect to the target current, b) A step of calculating the inverse Jacobian matrix, which is the inverse of the Jacobian matrix, c) A step of calculating a control value from the target current and the inverse Jacobian matrix, A triple active bridge converter, including...

5. A triple active bridge converter according to any one of claims 1 to 3, In step A) above, the control circuit calculates the control value by referring to a table that includes the correspondence between the target current, the detected port current, and the control value to be output. The aforementioned table is, a) A step of calculating the Jacobian matrix, which is an approximate model at each operating point, for the approximate current waveform of the port current with respect to the target current, b) A step of calculating the inverse Jacobian matrix, which is the inverse of the Jacobian matrix, c) A step of calculating a control value from the target current and the inverse Jacobian matrix, A triple active bridge converter created from results derived in advance under multiple conditions.

6. A triple active bridge converter according to any one of claims 1 to 3, The aforementioned inductance component is, Real element, Leakage inductance with the aforementioned transformer, or Actual elements and combination with the transformer This is a triple active bridge converter.

7. A multiactive bridge converter having three or more n ports, A transformer having n windings and a core that magnetically couples the n windings together, Each of the n windings is connected to n ports, A control circuit for switching the switching elements of each of the aforementioned ports, Equipped with, The aforementioned ports are as follows: A full bridge circuit including four switching elements, The inductance component directly connected to the aforementioned winding, Includes, The aforementioned control circuit is A) A process of calculating the control value using the approximate current waveform of the port current relative to the target current. Execute, The aforementioned approximate current waveform is derived from an approximate voltage waveform obtained by approximating the port voltage output from the full-bridge circuit with a continuous function. The aforementioned approximate voltage waveform is an approximate function obtained by superimposing the fundamental wave and harmonics onto the port voltage, which is a rectangular wave. Multi-active bridge converter.

Images